dissertation題目：business forcasting

dissertation語言：英語dissertation English

dissertation專業：F&A

字數：NO

學校國家：英國

是否有數據處理要求：是

您的學校：UCLAN

dissertation用于：BA Coursework 本科課程essay

截止日期：13/03/2100

補充要求和說明：

EXCEL MINITAB

 

Division of Systems and Operations

COURSEWORK QUESTIONS

 

Module No MG3003

Module Title Business Forecasting

Full Year

 

A  module of  20 credits                           Updated On 8/21/2013

Question_2

 

 

a) Explain what is meant by a 

 

i) a linear trend; and

ii) a quadratic trend, in time series modelling.

 

Hence discuss the level of differencing required to make such trends stationary.

 

(10 marks)

 

b) Explain what is meant by the following ideas in time series modelling.

 

i) auto-regressive component; ii) moving average component.

 

The general Box-Jenkins ARIMA(p,d,q) model can be written as: 

 

  

 

where d is the level\ of differencing; b(L) is a p-th order polynomial; and c(L) is a q-th order polynomial.

 

iii) Explain what this notation means.

 

Hence explain the structure of the following models for a time series yt.

 

iii) ARIMA(1,0,0) ; iv) ARIMA(0,1,1) , a0 = 0#p#分頁標題#e# v) ARIMA(1,2,1)

 

(40 marks)

 

b) Sales of a painkilling drug for a UK based pharmaceutical company have been collected as in CW_Sales_1011, which is available as an Excel file.

 

A deasonalised version of the pharmaceutical sales data is available in CW_Sales_Deseason_1011 as an Excel file. 

 

By building an ARIMA(1,1,1) model, based upon the deseasonalised sales data, forecast deseasonalised pharmaceutical sales for 2010. You should justify and interpret all key elements of your modelling. 

(50 marks)

 

Total: 100 marks